Guide to Cost PrediCtability in ConstruCtion

Guide to Cost Predictability in Construction:

An analysis of issues affecting the accuracy of construction cost estimates

Prepared by the Joint Federal Government / Industry Cost Predictability Taskforce

November 2012

Guide to Cost Predictability in Construction: An Analysis of Issues Affecting the Accuracy of Construction Cost Estimates November 2012

Table of Contents

Note from the Taskforce Chairman. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Taskforce Participation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Introduction to the Cost Predictability Issue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Issues and Considerations in Cost Predictability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Cost Estimate Variance Matrix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Recommendations for Improved Cost Predictability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Postscript . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Appendix 1 ? Classes of Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Appendix 2 ? Documents Recommended for Estimates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Appendix 3 ? Qualifications for Estimators and Cost Consultants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Appendix 4 ? Sources of Information and References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Enquiries should be directed to: Canadian Construction Association a-

This guide is the product of the individual taskforce members' collective input and does not purport to express or represent the position of their respective organizations or associations. It reflects recommended industry practices. Readers are cautioned that this guide does not deal with any specific fact situation or circumstance. This guide does not constitute legal or other professional advice. The taskforce and their respective organizations or associations do not accept any responsibility or liability for loss or damage which may be suffered as a result of its use or interpretation. Copyright 2012 Must not be copied in whole or in part without the written permission of the Joint Federal Government / Industry Cost Predictability Taskforce.

Guide to Cost Predictability in Construction: An Analysis of Issues Affecting the Accuracy of Construction Cost Estimates November 2012

Note from the Taskforce Chairman

What at first appeared to be a relatively simple task to review the various and mostly obvious causes of poor cost predictability increasingly grew in each taskforce meeting as more and more variables were raised and reviewed. As the topics for consideration ranged from first bid comparison, whether low, median or average, to cost growth during construction, to final project cost, and even whether to review various "soft costs" to the extent that they may have an effect on cost predictability, etc., it soon became apparent that a clear and rigid definition of the scope of the taskforce had to be established. This would limit the discussions to achieve definable recommendations relating to the original rationale for the taskforce, which was to analyze the increasingly large variances being experienced by PWGSC, DCC, and others between their "pre-tender estimate" and the subsequent contractors' bids.

The basic question whether such increasing variances are the result of the construction industry's "pricing strategies", or of an inaccurate or insufficient estimating process, was soon concluded to be the latter, since with few exceptions, under a normal and competitive bidding process actual bid prices are generally the best indicator of the prevailing competitive market conditions for a specific project at that time. Upon further review it was also found that the owners' estimating process was seldom properly completed up to a reasonable Class A level, and in most cases only up to Class B, or even a brought forward Class C!

Also, it must be recognized that each project is unique and different from the previous one in varying degrees. Although many projects may be relatively repetitive and "similar", such as "standard" office buildings or schools, many others are non-standard and individually complex, unique, one-of-a-kind, remotely located, etc., or any combination thereof, such as leading edge lab facilities, historic building renovations, environmentally challenged locations and increasingly green, sustainable, and more innovative structures and systems.

Further, since the preparation of project estimates has typically been based on historical information, such information may be outdated, or perhaps not even exist for certain individual types of projects; therefore, the

associated degree of accuracy of the estimate can be expected to vary accordingly.

For this reason the taskforce concluded that the previously accepted degree of accuracy for Class A estimates should be expanded from +/- 5% to a range of +/- 5 to 10%, to accommodate the range of such non-standard projects, with a similar variation applied to the other classes of estimates. As well, for projects involving significant renovations, or challenging geographic, environmental, or other unique circumstances, a further allowance should be analyzed and included.

In addition, whereas the previous cost predictability guidelines were related primarily to the degree of design development completion, it became obvious that with the increasing variety and complexity of various projects, the potential accuracy of estimates must equally recognize the relationship to the level of complexity of any particular project, or its level of estimating difficulty. For this reason, a Cost Estimate Variance Matrix was developed, to recognize the combination of both these dimensions in arriving at a reasonable estimate variance.

Finally, since estimating and bidding is not an exact science, the expectation for estimate accuracy must be realistic, in relation to the particular type of project and in proportion to the quality and time expended to produce a specific level of estimate accuracy. But at any level, adherence to the recommendations of this guide should definitely improve cost predictability.

In this respect it bears repeating the obvious, that to achieve a realistic Class A estimate requires the following ingredients:

1. Class A professional and experienced estimating qualifications;

2. Class A completed project design documentation; and

3. Class A time, sufficient to produce the required estimate analysis with accuracy.

Lastly, the taskforce was fortunate to have the benefit of senior, experienced, informed, articulate, and independent cross-sector representatives; I sincerely want to thank each of them for their effective and dedicated participation in the production of this guide.

1

Guide to Cost Predictability in Construction: An Analysis of Issues Affecting the Accuracy of Construction Cost Estimates November 2012

Executive Summary

Large discrepancies between pre-tender estimates and actual bids for construction have a serious impact on the viability of a project. Owners, architects, engineers, cost consultants, contractors and subcontractors all have a vested interest in ensuring a high degree of cost predictability. This issue, raised at the Federal Government / Industry Real Property Advisory Council, resulted in the formation of a Cost Predictability Taskforce, to research and recommend solutions to this increasing problem.

Defining Cost Predictability For the purpose of this guide, "cost predictability" is defined as "the prediction of a construction cost estimate, as compared to the median of competitive bids", where "prediction" is defined as "an assertion on the basis of data, theory or experience, but in advance of proof".

As such this guide will apply only to procurement processes that typically require the submission of a total project bid, such as a design-bid-build or design-build process, and although not directly applicable to the construction management process, it could also be applied to the sequential trade contractors' bids.

Similarly, many of the same underlying principles could also be applied to the further development of cost predictability guides for the increasing range of P3 and lease/purchase procurement process variations.

Taskforce Analysis There is no common statistical database on variances between pre-tender estimates and final bid results for the Canadian construction market. However, members of the task force shared internal data that showed up to 40% of tenders had low bids that varied, either up or down, by more than 30% from the pre-tender estimate and fewer than 20% of tenders had bids within 10% of the estimate. While this was not the case with all of the taskforce members, all agreed that cost predictability was an increasingly serious issue facing the industry.

It was agreed that the following key actions are required to ensure better cost predictability:

? engage qualified professionals to prepare estimates;

? use appropriate economic models in the estimating process;

? ensure that the project approval process recognizes cost predictability issues;

? include scope revision mechanisms in the project approval process; and

? recognize the degree of accuracy of the estimate being used.

A large number of factors contribute to variances in estimate to bid prices. This guide will provide practical advice and discuss the primary ways to improve the cost predictability of projects. The consequence of a failure is often a cancelled project.

The Cost Estimate Variance Matrix As shown in the Cost Estimate Variance Matrix, the accuracy of estimates varies throughout the project design cycle and according to the complexity of the specific project, and several other factors that may be unique to a project. Depending on the class of estimate and the complexity of the project, variances can range from 5% to 30%. If additional unique aspects or risks apply to a project, these variances should be analyzed and increased by an appropriate amount.

Improving Cost Predictability Early in the project life cycle, i.e., before the requirements are well-defined and before accurate estimates are available, owners will need to set budgets. The challenge for all stakeholders is to ensure that the appropriate scope and cost management framework is adopted to ensure the desired, or required, degree of cost predictability. The following recommendations should be considered to improve the cost predictability between the final pretender estimate and actual bid prices:

Before Tender ? include sufficient contingency to address market

volatility, timing of construction, and other exclusions in the estimate;

? consider possible scope variations in the tender as a contingency to adjust to the owner's budget;

? give designers sufficient time to finalize 100% bid documents for the pre-tender estimate; and

? allow cost consultants or estimators sufficient time to prepare and finalize pre-tender estimates.

During Tender ? keep the cost consultant involved during the tender

period;

? monitor addenda changes and bidding environment; and

? revise estimates to reflect scope changes and addenda.

After Tender ? involve the cost consultant in the post tender review;

2

Guide to Cost Predictability in Construction: An Analysis of Issues Affecting the Accuracy of Construction Cost Estimates November 2012

? analyze bid results against estimates for lessons learned;

? consider value engineering to improve bid versus estimate variations; and

? keep a record of historical estimate versus bid data.

Cost Predictability Guarantee There is no guarantee that pre-tender estimates will precisely match bid prices, just as there is no expectation that all bid prices will be the same. In reality, there are too many variables influencing final bid prices to expect 100% accuracy. However, following the recommendations in this guide will help improve and define the cost predictability of projects.

Taskforce Participation*

Chairman John Westeinde, P.Eng President Westeinde Construction Ltd.

Architect Alexander Rankin, OAA FRAIC ARIBA MRIAI FCDBI RCA Partner GRC Architects

Engineer Michael Allen, P.Eng President Adjeleian Allen Rubeli Limited

Cost Consultant Art Maw President Hanscomb Limited

Contractor Spencer Hagan, P.Eng GSC District Construction Risk Manager PCL Constructors Canada Inc.

Public Works and Government Services Canada Shawn Gardner Senior Director, AFD Directorate, Acquisitions Branch

Defence Construction Canada Ron deVries, P.Eng Senior Vice President, Operations

Michael Freemark Senior Technical Advisor

Canadian Museum of Nature Maria K. Somjen B.Arch. LL.B. (on leave from PWGSC) Project Director, VMMB Renewal Project

Canadian Construction Association Eric Lee Senior Director, Industry Practices

* NOTE: The positions and titles of the taskforce members are those which they held during the earlier development of this guide, and may have changed by the time it went to print.

3

Guide to Cost Predictability in Construction: An Analysis of Issues Affecting the Accuracy of Construction Cost Estimates November 2012

Introduction to the Cost Predictability Issue

The problem of cost predictability was introduced by Public Works and Government Services Canada (PWGSC) at the joint Government of Canada / Canadian Construction Association (CCA) meeting in April 2003, where the departments identified a need to improve the accuracy of its pre-tender construction estimates (Class A estimate) which were increasingly resulting in variances of greater than 30% from the bids received. This issue was delegated to the Project Delivery Working Group of the Federal Government / Industry Real Property Advisory Council for action and it was concluded that the problem of cost predictability was a generic problem, not only on federal government projects, but also for private industry projects. Therefore, it was subsequently decided to establish a separate taskforce to review this issue to:

? analyze the problems inherent in the cost predictability process;

? review procedures to improve cost predictability;

? develop a guide paper with recommendations to improve cost predictability; and

? develop procedures to monitor future cost predictability improvements over an initial five year period.

As a basic premise, the taskforce first noted that every construction project is unique and different from the last. Even if some buildings or projects may appear to be similar, circumstances such as varying site conditions, seasonal weather changes, different scheduling requirements, or local practices, can create significant estimating and cost differences. Even where buildings could be substantially similar, such as primary schools, there is still a tendency to redesign every next one for architectural diversity, technical advances, or other reasons.

In general, there is no easy way to establish consistent pretender estimates for even the most common construction projects that could form a reasonable market-specific and time-sensitive database for construction costs. While there are cost indices available, they are only useful in establishing utilitarian baselines, not in constructing an overall estimate that is within 5% of the median bid.

A complete estimate is a composite of literally thousands of individual estimate items and considerations, including construction techniques, innovative ideas, site supervision, productivity factors, front-end specifications, and scheduling, etc., all of which play crucial roles in arriving at

the final cost. As well, such factors as occasional mistakes (large or small), aggressive or conservative bidding for individual bidders' reasons, misinterpretations or misunderstandings of the bid documents, and various risk control procedures also play a significant role in the final outcome.

Bid closing involves an amalgam of prices being received by several individuals within each firm that must be evaluated and entered, the majority of which takes place within the last one-half hour period before bid closing, with major decisions happening up to the final minutes and seconds. Meanwhile, similar processes happen at the multi subcontractor and supplier levels.

In addition, local or national construction industries are becoming increasingly affected by global trends and influences, which can create significant and unexpected volatility and unpredictability in the construction marketplace. Recent examples of such market volatility would include:

? The volatility in the 2004 marketplace relating to structural steel and all steel-based commodity pricing.

? The abnormal cost escalation in BC during 2006 leading up to the Vancouver Olympics.

? The unprecedented run-up in almost all construction material prices in 2008, especially for steel and oil-based products, across the world.

? The equally unprecedented and sudden global financial crisis, and resulting potential temporary collapse of both the commodity and construction markets in 20082009.

All of this volatility requires an assessment of risk on the part of each and every contractor, subcontractor and sub-subcontractor who endeavour to strategize and plan for eventualities to establish a costing framework for the project duration, which could be many months or years, that fits within specific corporate risk tolerance levels, while at the same time striving to be the low bidder.

Therefore, as a result of this multi-variable bid process, it is not uncommon to see a range of variations in contractor bid prices. Generally this range can be up to 10%; however, for smaller and one of a kind projects this range can be as high as 50% or more.

The taskforce also found that many existing budgeting and project approval processes do not permit a sufficient degree of flexibility to deal with large variances in bid prices. This should be internally reviewed by each owner,

4

Guide to Cost Predictability in Construction: An Analysis of Issues Affecting the Accuracy of Construction Cost Estimates November 2012

in order to provide for sufficient organizational flexibility to be able to mitigate the results of such potential estimating variances and market volatility.

The taskforce also needed to consider whether the pretender cost estimates should be targeted for the eventual low contractor's bid, which often results in the actual contract award price, or whether it would be more prudent to target for the "median" of the various contractors' bids, which is generally considered to be closer to the "true" cost of the project.

Notwithstanding the obvious and necessary consideration attributed to the low bid value in any tendering scenario, there is a fundamental difficulty with attempting to target the low bid as the point of comparison in a cost predictability analysis. It is almost impossible to attempt to predict the effects of particularly innovative or unique bidding solutions on the part of individual bidders on any particular project because one or more may include: possible labour or specialized equipment advantages, locational or logistical advantages, a particularly aggressive bidding approach by anyone bidder ("this is my job" attitude), and even occasional mistakes etc. All of these items may combine to generate the low bid or to the contrary, if none of these competitive bidding solutions are applied, may result in a high bid.

In order to balance the effects of such bidding solutions with an understanding of the myriad of logistical elements,

market considerations, and costing variables inherent in the estimating process, the taskforce has selected the median bid approach as the most appropriate and professionally attainable for comparison of pre-tender estimates within this guide.

In summary, it must be recognized that given the myriad of variables that affect the contractor's bid at any given location and at any given time, it is highly problematic for any one individual or agency, no matter how qualified or "in tune" with local conditions and market influences, to arrive at a consistent and accurate pre-bid estimate for construction services. Perhaps the best that should reasonably be expected from such pre-bid estimates is a reasonable approximation of costs that reflect the many variables including the size, scope, complexity, and locality of a specific project.

Nevertheless, given the significant and apparently increasing variations between the bid and pre-tender estimate, it is necessary for the industry to analyze and improve on the various factors that affect cost predictability. Within this context, and based on an understanding of how this component of the industry operates, issues and considerations are recommended for review and guidance during the estimate preparation process.

5

Guide to Cost Predictability in Construction: An Analysis of Issues Affecting the Accuracy of Construction Cost Estimates November 2012

Issues and Considerations in Cost Predictability

1. Consideration of Local or Regional Market Influences

The preparation of pre-tender estimates should consider such specific factors as:

? the number of bidders on the project, and in each trade;

? the number of other projects being tendered at the same time;

? the likelihood that certain trades will only bid to certain contractors;

? complicated phasing of the project;

? pre-qualification of contractors;

? heritage issues;

? available time periods for tendering and construction;

? timing of the tender; and

? political mood.

The selection of a cost consultant should be reviewed to ensure previous experience with similar complexity and circumstances of the particular project, and to reflect a greater need for additional risk analysis specifically related to construction hard costs.

The pre-tender estimate should show the various influences that have been considered and the "factor" applied for these influences. A pre-tender estimate should clearly indicate what is "in" and what is "out", and why, so that some of the "out" items can be analyzed and added to a Risk Management Plan.

There is often a failure to recognize how a competitive marketplace, or lack thereof, may affect bid prices. Cost consultants should analyze market volatility and anticipated trends. Confirm with the cost consultant that they have consulted with local construction associations, as well as with local contractors and/or sub-contractors, as contractors who are regularly bidding on projects are best able to evaluate market volatility. Note that political decisions can also temporarily impact the local competitive market.

2. Current and Accurate Information Should Be Shared between Stakeholders

Obtain concurrence from all stakeholders that the same information is being used, otherwise stakeholders may use different numbers as reference points. During public construction projects several types of budgets

are produced and monitored in several ways. Therefore ensure that all stakeholders and government departments maintain the latest estimate provided by the cost consultant.

3. Cost Estimates Should Be Prepared during Tender Period for Changes to Schedule or Addenda

Request updates of the estimate during the bidding period. Ensure that each addendum is accompanied by a cost (and schedule) impact on the original estimate before the addendum is released. Ensure that the change is acceptable to the owner and not simply a "wish" list from the design team or a stakeholder. Keep in mind that schedule changes may also result in cost changes. This applies particularly to the Class A estimate stage.

4. Incomplete/Uncoordinated Bid Documents, Causing Different Interpretations

Improve the quality of tender documents and allow sufficient time for consultants to prepare and coordinate them. Complete a checklist for both constructability and coordination prior to tender. Addenda should be kept to a minimum, ideally with no more than three. Note that uncoordinated bid documents eventually result in change orders. Thus, past quality of tender documents could be used as criterion in the future performance evaluation of design consultants.

5. Failure to Recognize the Appropriate Class of Estimate

An earlier, less accurate, Class C or B estimate may sometimes be passed off as the pre-tender estimate, so ensure that a proper pre-tender estimate is prepared, developed from the 100% tender documents before tendering.

6. Fast -tracking Projects or Proceeding with a Project before It Is Well Defined

With a fixed end date, there is often a push to start work on a project long before all parameters are defined. This may cause a great deal of cycling, with the increased potential for incorrect estimates. For example, fasttracking may result in no pre-tender estimate. When it is determined that the estimates are wrong, the project is often well-advanced, and therefore difficult to realign or stop. Consequently, one should allow sufficient time for the preparation of estimates, or increase design contingencies using proper risk management analysis. Also, keep in mind that different procurement methods will have an impact on project estimates.

6

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download